Microbial elements as the initial triggers in the pathogenesis of polymorphic light eruption?

Abstract The primary trigger of polymorphic light eruption (PLE) remains to be uncovered. We hypothesize that PLE may be initiated by elements resulting from UV‐induced damage to microbial communities of the skin, leading to a cascade of events eventually resulting in the skin rash of the disease. One mechanism by which epidermal injury by UV radiation could trigger PLE are danger signals such as damage or pathogen associated molecular patterns DAMP/PAMPs or commensal‐associated molecular patterns (CAMPs). Such triggers could be produced due to UV‐induced stress on microbial communities of the skin and exacerbate inflammatory responses by inducing the innate immune system through antimicrobial peptides (AMPs) such as psoriasin, RNase7, HBD‐2 and LL‐37. These AMPs also actively take part in initiating adaptive immunity. That signals derived from microbial rather than human elements may initiate PLE is supported by series of observations, including the PLE‐protective effect of topically applied microbial‐derived DNA repair enzymes.


Abstract
The primary trigger of polymorphic light erup on (PLE) remains to be uncovered. We hypothesize that PLE may be ini ated by elements resul ng from UV-induced damage to microbial communi es of the skin, leading to a cascade of events eventually resul ng in the skin rash of the disease. One mechanism by which epidermal injury by UV radia on could trigger PLE are danger signals such as damage or pathogen associated molecular pa erns DAMP/PAMPs or commensalassociated molecular pa erns (CAMPs). Such triggers could be produced due to UV-induced stress on microbial communi es of the skin and exacerbate infl ammatory responses by inducing the innate immune system through an microbial pep des (AMPs) such as psoriasin, RNase7, HBD-2 and LL-37.
These AMPs also ac vely take part in ini a ng adap ve immunity. That signals derived from microbial rather than human elements may ini ate PLE is supported by series of observa ons, including the PLE-protec ve eff ect of topically applied microbial-derived DNA repair enzymes.

| BACKGROUND
As early as 1942, Epstein suggested a photoallergic concept for photodermatoses (S1). Since then, much progress has been made in understanding the pathophysiology of photodermatoses and their most common form, polymorphic light erup on ( PLE ) 1 (S2-S13). However, specifi c photoan gens that ini ate PLE have not yet been iden fi ed.
We hypothesize that PLE may be ini ated by elements resul ng from UV -induced damage to microbial communi es of the skin, leading to a cascade of events eventually resul ng in the skin rash of the disease.

| HYPOTHESIS AND PREMISES
One mechanism by which UV-R can trigger inflamma on are poten al danger signals, so-called damage or pathogen-associated molecular pa erns DAMP/PAMPs (S16, S17). These DAMP/PAMPs may exacerbate inflammatory responses by inducing the innate immune system, 2 for example by producing an microbial pep des (AMPs) 3 (Fig. 1 ), and in turn, these AMPs could contribute to the induc on of PLE. Similarly, damage to commensals may give rise to commensal-associated molecular pa erns (CAMPS) (S18, S19), resul ng in microbial signals, 2 altering the skin ' s microbial landscape and contribu ng to abnormal immune responses in inflammatory diseases such as psoriasis 4 and beyond. In this regard, studies indicate that photoprovoked PLE pa ents show an upregula on of certain AMPs 5 (S20). That signals derived from microbial rather than human elements may ini ate PLE is consistent with a series of observa ons. First, the fact that PLE is a very common condi on 6 requires that the puta ve trigger should be ubiquitous.
Undoubtedly, the skin ' s microbiome would fulfil this prerequisite. Second, PLE pa ents pretreated over 2 weeks twice daily with the topical vitamin D 3 analogue calcipotriol showed reduc on in PLE symptoms upon photoprovoca on. 7 Notably, it has been suggested that the beneficial effects of calcipotriol in inflammatory diseases are due to the modula on of AMP regula on via TH17 pathways. 5 In addi on, in vitro work in kera nocytes revealed that calcipotriol suppressed certain AMPs, which were s mulated by UVB. 8 On the other hand, vitamin D 3 has been found to upregulate LL-37 in normal skin (S21) and narrowband UVB irradia on increased serum LL-37 levels in psoriasis pa ents (S22). Indeed, direct effects of UV-R could affect AMP expression 3 and contribute in star ng the chain of events leading to PLE. Vitamin D is involved in modula ng the gut microbial communi es (S23), and downregula on of vitamin D receptor (VDR), associated with regula on of AMPs, is known to alter the microbial communi es and their func ons in murine intes ne (S24, S25) and may have an an microbial effect also for skin (S26). Furthermore, low vitamin D levels were reported in pa ents with PLE, most likely due to sunlight avoidance (S11, S27, S28).
Third, a lo on containing liposomal, unpurified microbial extracts with DNA repair capacity from Anacys s nidulans and Micrococcus luteus (a skin commensal) diminished PLE symptoms, without affec ng the physiologic sunburn (erythema) response. 9 The mechanism remained unclear; however, the direct effect of this enzyme approach on skin cells as measured by improved DNA repair was low (S29), and thus, we speculate that these extracts may have acted by repairing damage to commensals residing on or in the skin rather than reducing damage to human skin cells for which the approach has been intended for (S29, S30).

| HOW TO TEST THE HYPOTHESIS
First of all, it needs to be ascertained that the microbiome of the skin does not differ between PLE pa ents and the healthy controls by sequencing microbial DNA isolated, for example using skin swabs (S31). Tape stripping can be used to isolate and subsequently quan fy the skin ' s AMPs to study co-rela on between expression of AMPs and presence of microbial communi es. Furthermore, high-throughput sequencing of the adap ve T-cell immune receptor repertoire can be used to determine the clonality of an infiltrate (S32, S33). In this context, an immunologic type IV reacon to a specific photoan gen should be accompanied by a monoclonal T-cell infiltrate, whereas a reac on induced by mul ple an gens and/ F I G U R E 1 Hypothetical model for pathogenesis of PLE . Exposure of the skin to ultraviolet radiation ( UV -R) leads to the production of commensal-associated pattern ( CAMP ). They are usually buried intracellularly, but upon secretion from dying human cells or microbial communities of the skin, they may exacerbate inflammatory responses by inducing the innate immune system through producing antimicrobial peptides ( AMP s) such as S100A7 (psoriasin), HBD -2, RN ase7 and LL -37. This increase in AMP s can promote in a vicious circle the activation of adaptive immune responses and exacerbate the inflammatory responses. In addition, UV -R can also directly lead to microbial killing, resulting in the production of microbial signalling molecules such as lipopolysaccharides ( LPS ), lipoteichoic acid ( LTA ), oleic acid and others that in turn may lead to or enhanced abnormal immune responses through Toll-like receptor ( TLR ) activation and transcription factors such as NF -kB , AP -1 and IRF . In healthy subjects, various cytokines such as TNF , IL -4 and IL -10 are expressed and infiltration of neutrophils and Tregs and increase in mast cell numbers in the skin are observed upon UV exposure. However, in PLE , there is a reduced production of these cytokines and decreased infiltration of neutrophils and reduced numbers of mast cells in the skin. Moreover, Langerhans cell resistance to UV -R is seen in the skin of PLE patients compared to that of healthy controls. Taken together, all these events may be linked to the abrogation of UV -R-induced suppression of the adaptive immune responses in healthy subjects but a failure of suppression in PLE patients. Consequently, an influx of CD 4+ T cells and CD 8+ T cells is observed in the skin of PLE patients leading to inflammation and manifestation of the typical skin rash upon UV exposure | 1001 or CAMPS/DAMPS/PAMPS should be not. To directly study the role of the microbiome in the ini a on of PLE, experimental photoprovoca on could be performed in pa ents with and without rigid skin disinfec on (e.g. using topical chlorhexidine or povidone iodine) before repe ve UV exposure. Addi onally, in vitro grown cultures of skin commensals obtained from PLE pa ents could be exposed to UV-R and their extracts painted back onto human skin to inves gate whether PLE can be provoked. Ideally, a mouse model of PLE would be desirable to study the mechanisms of the disease but the only such model currently available has limita ons (S8). Finally, germ-free mice, an bio c-treated or disinfected mice could be used to study the effects of UV-R on the immune response by employing the contact hypersensi vity (CHS) model (S34).
A microbiome of a certain quality and quan ty could have a protec ve role against immune suppression and thus may contribute to reduc on of UV-induced immune suppression in PLE. For instance, downregula on of IL-10 was one cytokine abnormality found in PLE (S2) and a recent study suggested a protec ve role of commensal microbiota-derived IL-10 in stroke and inflammatory disorders (S19). Moreover, peripheral blood mononuclear cells from healthy donors when cocultured with probiotics produced large amounts of IL-10 (S35). Intriguingly, oral administraon of a nutri onal supplement containing lycopene, beta-carotene and Lactobacillus johnsonii ameliorated UVA-photoprovoked PLE (S36).

| RELEVANCE AND PERSPECTIVES
If the role of the microbiome in PLE is tested and confirmed, this will not only unravel the pathogenesis behind PLE, possibly open up avenues to new and be er therapeu c op ons, but also extend the knowledge to other areas such as the suscep bility to UV carcinogenesis. In this regard, a study by Lembo et al. (S37) and our previous work (S38) have suggested that suscep bility to PLE may be a protecve factor in the forma on of skin cancer. However, the ques on why PLE pa ents are resistant against UV-induced immune suppression (S13-S15) s ll remains open but gender (S39) and gene c predisposion may mainly account for this resistance (S40-S42).